The complexity of reachability in parametric Markov decision processes.
Sebastian JungesJoost-Pieter KatoenGuillermo A. PérezTobias WinklerPublished in: J. Comput. Syst. Sci. (2021)
Keyphrases
- markov decision processes
- state space
- optimal policy
- dynamic programming
- reinforcement learning
- policy iteration
- transition matrices
- finite state
- reachability analysis
- finite horizon
- planning under uncertainty
- decision processes
- average reward
- average cost
- reinforcement learning algorithms
- reward function
- decision theoretic planning
- markov decision process
- factored mdps
- action space
- model based reinforcement learning
- computational complexity
- partially observable
- infinite horizon
- heuristic search
- policy evaluation
- risk sensitive
- convergence rate
- state and action spaces
- discounted reward
- decision problems
- markov chain
- real time dynamic programming